Implication of Rituximab Infusion Reactions on Clinical Outcomes in Patients With Diffuse Large B-cell Lymphoma: A Single Institution Experience

Journal Pre-proof Implication of Rituximab Infusion Reactions on Clinical Outcomes in Patients with Diffuse Large B-Cell Lymphoma: A Single Institution Experience Dilan A. Patel, MD, Tanner M. Johanns, MD PhD, Kathryn Trinkaus, PhD, Nancy L. Bartlett, MD, Nina Wagner-Johnston, MD, Amanda F. Cashen, MD PII:

S2152-2650(19)31991-3

DOI:

https://doi.org/10.1016/j.clml.2019.09.604

Reference:

CLML 1421

To appear in:

Clinical Lymphoma, Myeloma and Leukemia

Received Date: 21 June 2019 Revised Date:

9 August 2019

Accepted Date: 21 September 2019

Please cite this article as: Patel DA, Johanns TM, Trinkaus K, Bartlett NL, Wagner-Johnston N, Cashen AF, Implication of Rituximab Infusion Reactions on Clinical Outcomes in Patients with Diffuse Large BCell Lymphoma: A Single Institution Experience, Clinical Lymphoma, Myeloma and Leukemia (2019), doi: https://doi.org/10.1016/j.clml.2019.09.604. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Elsevier Inc. All rights reserved.

1 Implication of Rituximab Infusion Reactions on Clinical Outcomes in Patients with Diffuse Large B-Cell Lymphoma: A Single Institution Experience Dilan A. Patel MDa, Tanner M. Johanns MD PhDa, Kathryn Trinkaus PhDc, Nancy L. Bartlett MDa, Nina Wagner-Johnston MDb, Amanda F. Cashen MDa a

Division of Medical Oncology, Department of Medicine, Washington University in St. Louis

School of Medicine, St. Louis, MO, U.S.A. b

Department of Oncology, Johns Hopkins University School of Medicine, Baltimore, MD,

U.S.A. c

Biostatistics Shared Resource, Siteman Cancer Center, St. Louis, MO, U.S.A

Correspondence: Amanda F. Cashen, MD Washington University School of Medicine 660 South Euclid Avenue St. Louis, MO 63108 Email: [email protected] Conflicts of Interest: The authors have no conflicts of interests to disclose.

2 Micro-abstract In a retrospective study of 229 patients with diffuse large B-cell lymphoma treated with rituximab-based chemotherapy regimens, we show that patients who experienced an infusion reaction had a higher probability of survival compared to those who did not. Abstract Background: The addition of the anti-CD20 monoclonal antibody rituximab to chemotherapy for diffuse large B-cell lymphoma (DLBCL) has led to improvements in progression-free survival and overall survival, although the exact mechanism of rituximab is not known. Rituximab administration often results in transient, non-life threatening infusion reactions (IR). We report a retrospective cohort of patients with DLBCL who received rituximab to determine the significance of IRs on clinical outcomes. Patients and Methods: We identified and analyzed a retrospective cohort of 229 patients with DLBCL. They were stratified into two cohorts, those who did and did not have an IR. Univariate and multivariate analyses were performed to evaluate the prognostic significance of rituximab-related IRs relative to DLBCL subtype, international prognostic index (IPI) score, cMyc translocations or amplifications, chemotherapy regimen, and Ki-67 proliferative index. Results: Baseline characteristics did not differ significantly between the two groups. Rituximab was included as initial treatment in all patients. Patients with an IR had a significantly higher overall survival (HR 0.26, 95% CI 0.07-0.95) at 5 years. In addition, subgroup analysis showed a significantly higher progression-free survival in patients with the germinal center subtype of disease and c-Myc alterations who had a rituximab-related IR (log-rank p<0.0001). Conclusions: The presence of a rituximab-related IR is associated with a better overall survival in patients with DLBCL. While limited by small sample size and retrospective nature, these results provide rationale for further investigation into the mechanism of action of rituximab in order to optimize the efficacy of CD20 monoclonal antibodies.

Keywords: immunochemotherapy, diffuse large B-cell lymphoma, rituximab, infusion reaction, complement

3 Introduction Rituximab is a chimeric mouse/human monoclonal antibody that binds to the CD20 receptor on both normal and neoplastic B-lymphocytes. Since being approved by the FDA in December 1997, rituximab has transformed medical practice for various non-hematologic and hematologic conditions, including indolent and aggressive subtypes of non-Hodgkin lymphoma (NHL)1, 2. The addition of rituximab to standard chemotherapy regimens such as cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) has led to improvements in progression-free survival and overall survival for patients with diffuse large B-cell lymphoma (DLBCL), the most common subtype of NHL3, 4, 5. The widespread success of rituximab has prompted the development of other CD20-targeting monoclonal antibodies2, 6, 7, 8, 9.

The primary mechanism by which rituximab mediates the elimination of CD20expressing lymphocytes is not known9. The two most widely accepted mechanisms of action are complement mediated cytotoxicity (CMC) and antibody dependent cell-mediated cytotoxicity (ADCC), both of which have evidence based on preclinical animal models10, 11. For example, studies in support of CMC have shown that mice deficient in various aspects of the complement system have less response to rituximab than mice with a fully intact innate immune system12, 13. Clinical studies have shown that the addition of fresh frozen plasma prior to rituximab infusion can augment an anti-tumor response13, 14, 15. ADCC, on the other hand, requires the Fc portion of the rituximab antibody to bind an Fc receptor (FcR)-expressing cell, namely NK cells and monocytes/macrophages, resulting in either direct cytolysis or phagocytosis, respectively16. Several mouse tumor models have shown impaired efficacy of rituximab in FcR-deficient mice12. Furthermore, polymorphisms in human FcR genes that result in enhanced binding, such as single nucleotide polymorphisms in Fc gamma receptor 2A and 3A, have been associated with longer progression-free survival following rituximab therapy16. Other pre-clinical studies, interestingly, have suggested that complement may impair ADCC, resulting in poorer anti-tumor responses17. Of note, type I monoclonal antibodies such as ofatumumab enhance CMC6, 8. Conversely, type II antibodies, such as tositumomab and obinutuzumab, weakly activate complement, but are thought to generate more ADCC6, 7, 8. These developments highlight the importance of gaining a deeper understanding of how these mechanisms contribute to the antilymphoma effect of CD20-targeted antibodies to help guide clinical utilization of these agents.

4

In general, rituximab is well-tolerated with minimal side effects. The most common complication is an infusion reaction (IR), which occurs in approximately one-third of patients and is typically limited to the first infusion. Clinical observational studies have suggested that complement can be activated during rituximab infusion, and the activation of complement is correlated with rituximab-associated infusion toxicity8, 18. Given the conflicting role of CMC in either directly mediating an anti-lymphoma effect or impairing ADCC-mediated anti-lymphoma immunity, we asked if the presence or absence of a rituximab-mediated IR is associated with differential outcomes in patients with DLBCL8, 18. These results will provide rationale for future studies utilizing novel CD20 monoclonal antibodies with different CMC-activating potential in DLBCL.

Methods Patients We determined the characteristics and outcomes of 229 patients with DLBCL who received their initial treatment with a rituximab-containing chemotherapy regimen at Siteman Cancer Center at Washington University in St. Louis School of Medicine. Institutional Review Board (IRB) approval was obtained prior to the initiation of the study.

Inclusion criteria included patients greater than 18 years of age with histologically proven DLBCL who received their first infusion of rituximab between January 2000 and June 2015 at Siteman Cancer Center and had at least 6 months of follow-up treatment records. Patients with germinal center B-cell (GCB), non-GCB, primary mediastinal B-cell (PMBCL), and double-hit subtypes of DLBCL were included in the study. DLBCL subtype was determined by immunohistochemistry for 180 patients in our study20. FISH analysis was performed on 226 patients for determination of C-MYC rearrangements or amplifications.

Exclusion criteria included patients who either did not receive rituximab or who received the first infusion at another institution or without appropriate documentation of the first infusion. Patients with a history of low grade lymphoma of chronic lymphocytic leukemia were not

5 included. Additionally, patients with primary CNS lymphoma (PCNSL) or HIV-associated DLBCL were excluded.

The Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 was used to grade the severity of adverse drug events (ADE) based on review of physician and nursing notes in the medical record. Only grade 2 or greater IRs were considered. OS was defined as time from treatment to death from any cause. PFS was defined as time from initial treatment to radiographic or biopsy proven relapse, progression, or death. Follow-up data were obtained through June 2015.

Statistical Analysis The Jonckheere-Terpstra test was used for ordinal variables such as international prognostic index (IPI) score and stage (Table 1). Chi-square tests were used to determine pvalues of GCB, c-Myc, IPI (low versus high) and stage (Table 1). The p-value for Ki-67 was calculated from a Wilcoxon rank-sum test. Cox proportional hazards models were used to calculate overall survival (OS), adjusted for c-Myc status and IPI score, along with hazard ratios (HR), 95% confidence intervals (CI), and associated p-values. All statistical tests were twosided, with p-value <0.05 used to indicate statistical significance.

Results Patient Characteristics Table 1 shows the baseline and pre-treatment patient and disease characteristics for patients who did and did not experience an IR. Only the choice of chemotherapy regimen (RCHOP versus other) and c-Myc alterations (rearrangement or amplification) statistically differed between the two groups (Table 1). Patients in both groups, however, were similar in terms of IPI score (composite of age, stage, performance status, serum lactate dehydrogenase, and number of extra-nodal sites), DLBCL subtype, stage, and Ki-67 proliferative index. The most common chemotherapy regimen used was R-CHOP in 70% of patients. Most of the patients who did not receive R-CHOP had higher risk disease, such as those with c-Myc mutations, and therefore received more intense regimens such as dose adjusted R-EPOCH (rituximab, etoposide,

6 prednisone, vincristine, cyclophosphamide, doxorubicin) or R-ICE (rituximab, ifosfamide, cyclophosphamide, etoposide),

Impact of Infusion Reaction on Survival To determine the association of IR on clinical outcome, we assessed the impact of the presence or absence of a rituximab-related IR on survival. Patients with an IR had a statistically significant increase in OS at 5 years (Table 2, HR 1.35, p-value = 0.04). In evaluating the effect of IR on overall survival, the presence of an IR was associated with a 74% lower risk of death at 5 years (HR 0.26; CI 0.07-0.95), suggesting that the presence of a rituximab-associated IR portends a favorable prognosis. Consistent with previous data, univariate analysis also demonstrated a good prognostic association with presence of GCB subtype and poor prognostic association with presence of c-Myc alterations (Table 2).

Due to the association between c-Myc alteration and IPI score with the presence of IR, overall survival data was adjusted for c-Myc status and IPI score (Figure 1). The presence of an IR was significantly associated with improved OS (HR 0.002, 95% CI 0.00 to 0.41, p-value = 0.02). In subgroup analyses, a strong protective effect was noted at IPI scores of 0-3 with presence of an IR (Table 3). At IPI scores of 4-5, there was a non-significant difference in OS regardless of the presence or absence of an IR (Table 3). Overall, the presence of IR was associated with improved OS and this was most pronounced in patients with lower IPI scores.

In subgroup analyses, c-Myc altered GCB patients with IR had a statistically significant improved progression-free survival (PFS) (Figure 2, log-rank p-value <0.0001). Consistent with other studies, our data showed that C-Myc alterations were associated with the GCB subtype (18% versus 3%)21, 22. The patients with the worst PFS had c-Myc alterations without an IR, with eight of ten patients either dying or having progressive disease (Figure 2). In multivariate analyses, no significant differences in PFS were seen for other subgroups, including the entire cohort, patients with non-GCB or GCB subtypes alone, or patients with the GCB subtype without c-Myc alterations, suggesting that the benefit of an IR occurs almost exclusively in a particular group of patients with the GCB subtype and c-Myc alterations (Figure 2). No demonstrable association was found between disease type, either GCB or non-GCB, IR, or c-

7 Myc alterations, suggesting that the association of the GCB subtype with c-Myc alterations is unlikely due to confounding (Table 4).

Discussion Despite the frequency of rituximab-related IRs, little is known about the implications of such reactions on outcomes or which patients are most likely to develop a reaction. Our retrospective study of patients with de novo DLBCL treated with rituximab-based chemotherapy regimens showed that patients who had an IR had a significantly prolonged OS. As expected, OS also correlated with known baseline patient and disease characteristics, as reflected by the IPI score and GCB status, with the latter reflective of more favorable outcomes compared to the nonGCB subtype27. In our study, patients with lower IPI scores were more likely to develop an IR and to have better outcomes when an IR occurred with the first infusion of rituximab, a finding that has not been described previously. Patients with higher IPI scores had worse OS regardless of the presence or absence of an IR, suggesting that a potential beneficial effect from an IR is negated by aggressive underlying disease biology.

The multivariate analysis association of c-Myc positive disease and risk of IR was unexpected, given the more aggressive nature of such lymphomas21, 22. Although limited by a smaller sample size, our data indicate a higher risk of developing an IR among this subgroup, though the occurrence of an IR did not predict for a higher OS, consistent with the observation in higher IPI score patients where more aggressive nature negates the potential beneficial effect associated with an IR21, 22. These patients had an approximately six-fold higher hazard of death compared to patients without rearrangements or amplifications, consistent with previous reports, and suggestive that there is nothing intrinsically different about the cohort included in this study21, 22, 23, 24, 25. Furthermore, the association of c-Myc with survival was independent of the presence or absence of IRs or the IPI score, in part because of the close relationship between cMyc alterations and GCB subtype. Interestingly, the subgroup analysis finding that the progression-free survival benefit for IR exists almost entirely in the patients with GCB subtype and c-Myc alterations is novel, suggesting that these patients should be included in future prospective clinical trials to optimize the benefit of monoclonal antibody therapy in this traditionally high-risk group (Table 2).

8

Together, given the proposed involvement of complement in the underlying pathophysiology of rituximab-associated IRs and our data suggesting the potential protective association of IRs in a subgroup of DLBCL patients, these results suggest that future studies aimed at augmenting the CMC-associated response to rituximab may be warranted. Indeed, it is intriguing to speculate that perhaps CMC-augmenting strategies such as pre-rituximab infusion of fresh frozen plasma or the use of type I antibodies such as ofatumumab, which enhance CMC, may be potentially more effective in certain cohorts of patients26, 27, 28. Lastly, gaining a deeper mechanistic understanding of the role of CMC in rituximab-mediated clearance is also necessary.

It should be noted our study has several limitations. Firstly, this is a retrospective, single institution study with a limited number of patients and events. Secondly, our observations do not provide information regarding the underlying mechanism of rituximab activity or the nature of IRs. As a result, larger cohorts of patients are needed to validate these findings and to determine if such observations have any therapeutic relevance on outcomes for patients with DLBCL, particularly for the subgroup of GCB patients with c-Myc alterations. Lastly, it is unclear if the observations described herein are specific to DLBCL or are also applicable to other leukemias or lymphomas treated with CD20-targeted antibodies, such as follicular lymphoma or CLL.

Conclusions In summary, we observed that rituximab-related IRs are associated with significantly prolonged overall survival in patients with DLBCL. These results support the hypothesis that CMC is important for the efficacy of rituximab and that strategies to augment CMC may have therapeutic implications.

Clinical Practice Points 1. Although rituximab-related infusion reactions (IRs) are a relatively common phenomenon occurring in approximately one-third of patients and often self-limiting, the exact mechanism behind the phenomenon is unclear, though complement-mediated cytotoxicity (CMC) and antibody-dependent cell-mediated cytotoxicity (ADCC) are thought to be important.

9 2. Patients with DLBCL who had an IR with initiation of chemoimmunotherapy had a higher overall survival probability at 5 years, particularly those with lower IPI scores. 3. These data suggest that the mechanism behind IRs should be studied further so that future monoclonal antibody therapy can be tailored to provide optimal therapeutic benefit.

Author Contributions DP and TJ collected data. KT and TJ analyzed the data. DP and TJ wrote the manuscript. DP, TJ, NB, AC, and NWJ developed the conceptual framework for the study and reviewed and approved the manuscript prior to submission. References 1. Dotan E, Aggarwal C, Smith MR. Impact of rituximab (Rituxan) on the treatment of B-cell non-Hodgkin’s lymphoma. P T. 2010; 35: 148-157. 2. Salles G, Barrett M, Foa R, Maurer J, O’Brien S, Valente N, Wenger M, Maloney D. Rituximab in B-Cell Hematologic Malignancies: A Review of 20 Years of Clinial Experience. Adv Ther. 2017; 34: 2232-2273. 3. Vose JM, Link BK, Grossbard ML, Czuczman M, Grillo-Lopez A, Gillman P, Lowe A, Kunkel LA, Fisher RI. Phase II study of rituximab in combination with CHOP chemotherapy in patients with previously untreated, aggressive non-Hodgkin’s lymphoma. J Clin Oncol. 2001; 19: 389-397. 4. Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large B-cell lymphoma. N Engl J Med. 2002; 346: 235-242. 5. Habermann TM, Weller EA, Morrison VA, et al. Rituximab-CHOP versus CHOP alone or with maintenance rituximab in older patients with diffuse large B-cell lymphoma. J Clin Oncol. 2006; 24: 3121-3127. 6. Bauer K, Rancea M, Roloff V, Elter T, Hallek M, Engert A, Skoetz N. Rituximab, ofatumumab, and other monoclonal anti-CD20 antibodies for chronic lymphocytic leukaemia. Cochrane Database Syst Rev. 2012; 11: CD008079. 7. Tobinai K, Klein C, Oya N, Fingerie-Rowson G. A review of obinutuzumab (GA101), a novel type II anti-CD20 monoclonal antibody, for the treatment of patients with B-cell malignancies. Adv Ther. 2017; 34: 324-56.

10 8. Alduaij W, Illidge T. The future of anti-CD20 monoclonal antibodies: are we making progress? Blood. 2011; 117: 2993-3001. 9. Maloney DG, Smith B, Rose A. Rituximab: mechanism of action and resistance. Semin Oncol. 2002; 29 (1 Suppl 2): 2-9. 10. Boross P, Leusen JHW. Mechanisms of action of CD20 antibodies. Am J Cancer Res. 2012; 2: 676-690. 11. Weiner GJ. Rituximab: mechanism of action. Semin Hematol. 2010; 47: 115-123. 12. Uchida J, Hamaguchi Y, Oliver JA, Ravetch JV, Poe JC, Haas KM, Tedder TF. The innate mononuclear phagocyte network depletes B lymphocytes through Fc receptor-dependent mechanisms during anti-CD20 antibody immunotherapy. J Exp Med. 2004; 199: 1659-1669. 13. Di Gaetano, N, Cittera E, Nota R, Vecchi A, Grieco V, Scanziani E, Botto M, Introna M, Golay J. Complement activation determines the therapeutic activity of rituximab in vivo. J Immunol. 2003; 171: 1581-1587. 14. Klepfish A, Rachmilewitz EA, Kotsianidis I, Patchenko P, Schattner A. Adding fresh frozen plasma to rituximab for the treatment of patients with refractory advanced CLL. QJM. 2008; 101: 737-740. 15. Klepfish A, Gilles L, Ioannis K, Rachmilewitz EA, Schattner A. Enhancing the action of rituximab in chronic lymphocytic leukemia by adding fresh frozen plasma: complement/rituximab interactions & clinical results in refractory CLL. Ann N Y Acad Sci. 2009; 1173: 867-873. 16. Weng WK, Levy R. Two immunoglobulin G fragment C receptor polymorphisms independently predict response to rituximab in patients with follicular lymphoma. J Clin Oncol. 2003; 21: 3940-3947. 17. Wang SY, Racila E, Taylor RP and Weiner GJ. NK-cell activation and antibody-dependent cellular cytotoxicity induced by rituximab-coated target cells is inhibited by the C3b component of complement. Blood. 2008; 111: 1456-1463. 18. van der Kolk LE, Grillo-Lopez AJ, Baars JW, Hack CE, van Oers MH. Complement activation plays a key role in the side-effects of rituximab treatment. Br J Haematol 2001;115(4):807–811. 19. International Non-Hodgkin’s Lymphoma Prognostic Factors Project. A predictive model for aggressive non-Hodgkin’s lymphoma. N Engl J Med. 1993; 30(14): 987-994.

11 20. Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004; 103(1): 275-282. 21. Cabanillas F, Shah B. Advances in Diagnosis and Management of Diffuse Large B-cell Lymphoma. Clin Lymphoma Myeloma Leuk. 2017; 17(12): 783-796. 22. Petrich AM, Nabhan C, Smith SM. Myc-associated and double-hit lymphomas: a review of pathobiology, prognosis, and therapeutic approaches. Cancer. 2014; 120(24): 3884-3895. 23. Quesada AE, Medeiros LJ, Desai PA, et al. Increased MYC copy number is an independent prognostic factor in patients with diffuse large B-cell lymphoma. Mod Pathol. 2017. 24. Zhou K, Xu D, Cao Y, Wang J, Yang Y, Huang M. C-MYC aberrations as prognostic factors in diffuse large B-cell lymphoma: a meta-analysis of epidemiological studies. PLos One. 2014; 9(4): 1-9. 25. Perry AM, Alvarado-Bernal Y, Laurini JA, et al. MYC and BCL2 protein expression predicts survival in patients with diffuse large B-cell lymphoma treated with rituximab. Br J Haematol. 2014; 165(3): 382-391. 26. Nowakowski GS, Czuczman MS. ABC, BCB, and Double-Hit Diffuse Large B-Cell Lymphoma: Does Subtype Make a Difference in Therapy Selection? Am Soc Clin Oncol Educ Book. 2015: 449-457. 27. Gleeson M, et al. The Activated B-Cell Subtype of Diffuse Large B-Cell Lymphoma As Determined by Whole Genome Expression Profiling on Paraffin Embedded Tissue is Independently Associated with Reduced Overall and Progression Free Survival in the Rituximab Era: Results from the UK NCRI R-CHOP 14 v 21 Phase III Trial. Blood. 2016; 128: 1746 Shipp et al. A Predictive Model for Aggressive Non-Hodgkin’s Lymphoma. N Engl J Med. 1993; 329(14):987-994. 28. Sopp J and Cragg S. Deleting Malignant B Cells With Second-Generation Anti-CD20 Antibodies. J Clin Oncol. 2018; 22(1): 2323-2325.

Table 1: Baseline Characteristics

Variable DLBCL Type GCB non-GCB c-Myc alteration No Yes Chemotherapy R-CHOP Other IPI Score 0-3 4-5 Stage I II III IV Ki-67 (median and IQR)

Rituximab infusion reaction (IR), (0-1 = no reaction, ≥2 = reaction) 0-1 ≥2 73 (58%) 53 (42%)

35 (66%) 19 (35%)

144 (93%) 11 (7%)

56 (82%) 12 (18%)

p-value 0.39

0.02

0.04 113 (72%) 45 (29%)

39 (57%) 29 (43%) 0.77

94 (77%) 28 (23%)

39 (75%) 13 (25%)

18 (12%) 32 (20%) 17(11%) 90 (57%) 70 (45-80)

10 (15%) 12 (18%) 7 (10%) 38 (57%) 60 (4580)

0.90

0.26

Table legend: IR, infusion reaction; DLBCL, diffuse large B-cell lymphoma; GCB, germinal center B-cell; ABC, activated B-cell subtype; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone; other, chemotherapy regimens that are not R-CHOP; IPI, International Prognostic Index; IQR, Interquartile Range

Table 2: Univariate Analysis of Infusion Reactions (IR) as a Factor on Overall Survival (OS) Hazard Standard Ratio Error Chi-Square

Parameter

p-value

Infusion Reaction (Yes vs. No)

1.35

0.66

4.12

0.04

GCB TYPE (GCB vs. Non-GCB)

1.34

0.66

4.18

0.04

C-Myc Amplification or Rearrangement (Yes vs. No)

2.80

1.11

6.39

0.01

Regimen (DA-EPOCH-R vs. R-CHOP)

0.62

1.09

0.33

0.56

Regimen (Other vs. R-CHOP)

0.78

0.57

1.89

0.17

Ki-67 (0-100%)

0.00

0.01

0.01

0.93

IPI Score (0-5)

0.56

0.22

6.63

0.01

Table legend: GCB; germinal center B-cell like, Non-GCB, non-germinal center B-cell like subtype, DA-EPOCH-R, dose-adjusted etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, with rituximab; R-CHOP, rituximab with cyclophosphamide, doxorubicin, vincristine, prednisone; IPI score, International Prognostic Index Table 3: Hazard Ratios for overall survival at Each Level of IPI Hazard Ratios for IR at each level of IPI Description

Hazard Ratio

95% Wald Confidence Limits

IR (Yes vs. No) At IPI=0

0.000

0.000

0.116

IR (Yes vs. No) At IPI=1

0.000

0.000

0.196

IR (Yes vs. No) At IPI=2

0.001

0.000

0.335

IR (Yes vs. No) At IPI=3

0.012

0.000

0.594

IR (Yes vs. No) At IPI=4

0.176

0.025

1.22

IR) (Yes vs. No) At IPI=5

2.56

0.675

9.67

Table legend: IR, infusion reaction; IPI, International Prognostic Index

Table 4: Multivariate Analysis of Factors Associated with Risk of Developing an Infusion Reaction

Parameter

Hazard Ratio

95% Wald Confidence Limits

p-Value

Infusion reaction (IR) (Yes vs. No)

0.002

0.000

0.409

0.018

C-Myc Amplification or Rearrangement (Yes vs. No)

5.80

2.58

13.08

<.0001

IPI Score (0-5) without IR

1.42

1.06

1.89

0.016

IPI Score (0-5) with IR

20.58

2.39

176.95

0.016

Table legend: IR, infusion reaction; IPI, International Prognostic Index

Figure 1: Survival of Patients with (Red) and without (Blue) Infusion Reactions (IRs)

Figure legend: Includes data from all patients, adjusted for c-Myc aberration status, and IPI (when known) to account for established prognostic impact of these factors IR, infusion reaction; IPI, International Prognostic Index

Figure 2: Profession-Free Survival by IR and C-Myc status in GCB Patients

Figure legend: Infusion reaction, IR; germinal center B-cells, GCB